management decision theory

ManagementDecision TheoryManaging Pragmatically - 2nd editionBy Howard Flomberg

Included in this preview:

• Copyright Page

• Table of Contents

• Excerpt of Chapter 1

For additional information on adopting this book for your class, please contact us at 800.200.3908 x71 or via e-mail at [email protected]

Sneak Preview

Management Decision Theory Second Edition

by Howard Flomberg

Copyright © 2008 by Howard Flomberg.No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without permission in writing from the publisher. University Readers is NOT affiliated or endorsed by any university or institution.

First published in the United States of America in 2008 by University Readers

Cover Concept by Fabián ChowCover design by Monica Hui Hekman

12 11 10 09 08 1 2 3 4 5

Printed in the United States of America

ISBN: 978-1-934269-29-9 (paper)

Contents

Acknowledgements 1

SeCtIoN oNe: AN oveRvIew 3

Flomberg’s Laws 5Introduction 7the Decision Model 11operations Research 13Problem Definition 14Case Studies 16

SeCtIoN two: QUANtItAtIve ANALySIS 18 the Basics: Statistics and Probability 21Central tendencies 22 Mean and Median 22 Standard Deviation 23examples: Statistics Gone wrong 25 example 1: Stock Prices 25 example 2: Gosdork Inc. 27 example 3: Calculating Percentages 28 Example 4: Quantifiable

vs. Qualifiable variables 28Probability 29

Bayes’ Rule (Conditional Probabilities) 30 Bayes’ theorem 32 the tabular Approach 33Study Questions 35Utility theory 37 Risk Premium 39Game theory 40 Pascal’s wager 41 the Prisoner’s Dilemma 43 Zero-sum Games 44 the Nash equilibrium 46 Dominant Strategies 48 Game theory 1 50 Game theory 2 51 the Bottled water Problem 51 the two Farmers 52 More Games 53 Study Questions 54value of Information 55 expected value of Perfect Information (evPI) 55 expected opportunity (eoL) 57 Study Questions 59Analytic Hierarchy 60 Problem: what College Shall I Attend? 60 Pairwise Comparisons 62 Study Questions 64Forecasting 65 trend Analysis 65 Moving Averages 67 the Realities of Forecasting 68

Linear Programming 68 Formulating the Problem 69 the Graphic Solution 70 Linear Regression 74Regression Analysis 81 Regression Analysis 2 82 Regression Analysis 3 82 Correlation 83–84Case Study: Roybal Courier Service 85Fuzzy Logic and the Power of Iteration 89Bridging the Gap— Combining Quantitative and Qualitative techniques (Is that Possible?) 90 Incompatibility thesis 90ethical Pragmatism 91 the Principle of Utility: John Stuart Mill 92 Negative Utilitarianism 92 Preference Utilitarianism 93 ethics 93 objectivism—Ayn Rand 93 Judeo–Christian view 94

SeCtIoN tHRee: QUALItAtIve ANALySIS 95

Prototyping 99Prepare an Agenda 100Brainstorming and the Straw Man 102the Jury of executive opinion 103the Delphi Method 104 Step 1: Selecting the Panel of experts 106

Step 2: Questionnaire #1 107 Step 3: Questionnaires #2 through n 108 Step 4: evaluation 109example: Flomberg State College extended Campus 110the theory of Constraints (toC) 113 Implementing toC 114 Applying toC 114Case Studies 118 MasterCard International 118 Second Survey 120 Strategic Petroleum Reserves 123 education in the USA 123F & L University 124 Scenario 124University of Alberta Increases timely Access to Policies and Procedures 125 Background 125 How the University of Alberta’s Internet Site works 126 Benefits of Using Stellent Universal Content Management 127 Stellent Universal Content and Process Management Features 128Another Look at wal-Mart 129

SeCtIoN FoUR: PUttING It ALL toGetHeR: DeveLoPING tHe APPRoPRIAte eNvIRoNMeNt 133

Making the Final Decision 135Project (tASK) Management 139 overview 139 the Hodge Construction Company 141 Loading People into the Model 149

Bibliography 151

Acknowledgments 1

Acknowledgments

P eople who are mathematically inclined are frequently semi-literate. With that in mind, I would like to thank my son, Dave Flomberg for his support and encouragement, and for his assistance in turning my feeble

attempt at communication into acceptable English. Most of the editing for the first edition was squeezed into his unbelievably tight schedule. Thanks Dave!

We are all products of our education. I would like to include one of my teachers and mentors Dr. Joe Megeath for giving me an invaluable education on real values.

One of the ongoing lessons that I receive is that I frequently learn more from my students than I realize. A paper turned in by a student, Melanie Tyler, brought many of the points that I was trying to tie up into focus. For that I thank her. This is yet another reminder that a good teacher frequently learns from the student.

I’d also like to thank Brooke Pruter for her insightful feedback and perspective.

This second edition has given me a chance to improve the book. I was able to fix some arithmetic errors, tweak the text and add come content.

There is a new section Analytic Hierarchy. Some faculty from other schools requested this tool. I am also putting many of the tables into Excel© spreadsheets. The publisher will make these available to instructors.

2 Management Decision Theory

I’d like to thank the folks at University Readers. Their enthusiasm and pro-fessional support made the project flow. I heartily recommend them to other professors, especially those neophyte writers out there.

Ethical Pragmatism: “Doing what works best” while mindful of the obliga-tion one has to society.

Howard P. FlombergDenver Colorado

SectionOne

An Overview

Section One 5

Flomberg’s Laws

There are no replacements for the ultimate managerial tool, “Common 1. Sense.” 1

You can’t make a decision if you don’t define the problem. 2. Doing nothing is always a valid alternative. 3. Sometimes the best answer is, “It Depends.” 4. Be very careful when quoting statistics. They can be very misleading. 5. The word “Accurate” is by definition a very inaccurate term. 6. Use the tool that makes the most sense. 7. If you do not have an effective organization, nothing will work well.8.

1com·mon sense (n)—Sound practical judgment derived from experience rather than study.

Section One 7

Introduction

D ecision-making has been a black art for centuries. In the 20th Century, however, methods and procedures for decision-making have achieved some success, thanks to management science techniques.

Making a decision is, by its very nature, a blend of qualitative and quantitative processes.

Qualitative analysis is built around scrutiny of observed or anticipated ac-tions. This research technique demands an analyst who can maintain an objec-tive view of the situation. However, when we discuss quantitative analysis, we think of numbers and quantities. The mind wanders to counting, statistics and probabilities, an uncomfortable place for many. This has been the standard domain for decision theory for decades.

Statisticians and the mathematically inclined consider qualitative analysis to be a stepchild. In contrast, a person who is involved in the decision making process often intuitively operates using qualitative analysis. Qualitative analysis makes use of that person’s experience, expertise and professional opinions.

This study revolves around techniques that use both qualitative and quantita-tive approaches. Some of the tools that will be covered are: Bayes Theorem, Game Theory and The Delphi Method.

Bayes Theorem evaluates probabilities with the assumption that past events can affect future events. Bayes gives us a way of using history to predict the


future. Bayes is frequently taught using mathematic notation. Most non-math-ematicians are inexperienced or uncomfortable with this notation. Alternative methods include tree theory and table manipulation. In this text, Bayes will be presented for the less-mathematically oriented reader.

John Von Neumann, the great Hungarian–American mathematician, co-wrote the classic thesis on Game Theory, Theory of Games and Economic Behavior (1944). Another great Princeton University mathematician, John F. Nash2, fur-ther advanced Game Theory. Game theory gives the ability to methodologically examine a decision and evaluate the optimal payoffs and penalties. 3

The Delphi method is a formal way of gathering the appropriate people so that a decision can be made. The technique comes from work done at the Rand Corporation in the 1950s. The technique was used to model of the effects of nuclear war. It traces directly back to the operations research work done by the British during World War II. Military planners have used decision theory very successfully. In his recent book “blink4” Malcolm Gladwell makes a point of comparing the similarities between Wall Street types and military planners. The point being made is that the frenetic world of the stock market as well as the complexities of conducting a military campaign both shares the need for immediate and effective decisions. This text will also evaluate and capitalize on Mr. Gladwell’s concepts.

Decision-making is enabled by sound management principles. A well-run organization is an environment where appropriate decisions will be made. A poorly run organization frequently forces poor decisions. This book is essen-tially a toolbox that provides both qualitative and quantitative tools that will aide in decision-making. However if the organization does not work, nothing else will.

Probably the first recorded instance in identifying a management methodol-ogy is in the Bible. Exodus 18:12–27 tells the story of Moses’ being totally swamped settling the daily problems of the Israelites. The passage tells us

2 Nash was the subject of the recent book (1998, Sylvia Nasar, Simon and Schuster Publishing) and subsequent movie “A Beautiful Mind.” (2001, Universal) 3At one point in time, Nash, Von Neumann and Einstein were all resident at Princeton University—WOW4Gladwell, Malcolm. Blink. New York: Little Brown and Company, 2005

Section One 9

that Jethro, Moses’ father-in-law comes to the camp bringing Moses’ wife and children.

Apparently Jethro is quite taken aback at the lack of organization and Exodus 18:21 Jethro tells Moses:

“…thou shalt select out of all of the people able men, such as fear G–d. Men of truth, hating bribes; and place these over them as officers over thousands, officers over hundreds, officers over fifties and officers over tens.”

Having given Moses this advice, like any good consultant, Jethro goes home to Midian. The key point is that Jethro tells Moses to establish a hierarchical organization. Identify leaders at the strategic, tactical and operational levels and have the decision made at the lowest possible level. This is good advice today. Note: I am quoting from the Bible as a historic reference only.

As I said, a good organization provides a bed for good decisions.

Section One 11

The Decision Model

In high school, we learned the “Scientific Method.”

Question

Observe

Formulate Hypothesis

Decide

Predict Results

Review Results


Business has converted this process to what can we will be calling the Decision Analysis model. This is the model we will use for formal decision modeling:

Define the Problem Identify

Alternative Solutions

Identify future Outcomes

Evaluate outcomes, Positive or Negative

Decide

Review Results

Section One 13

oPeRAtIoNS ReSeARCH

Operations Research can be considered the godfather of decision analysis. Its roots go back to World War II. Given the vastness of the logistics of running the war, British (and later, American) military management called upon the scien-tific community to produce a method to solve immense problems. The resulting method, operations research, involved gathering mixed teams from multiple disciplines to address these problems. The tools developed by this approach were primarily quantitative. The approach to decision making, however, was revolutionary, as the man said, “The Medium is the message.5”

Operations research has been adapted by business and science to solve many problematic situations. However, the technique of using mixed teams of profes-sionals to attack a problem has become a major force.

Operations research places a heavy reliance on mathematic modeling. The methods include:

Linear Programming • Network Analysis—Pert, Gantt Charts, etc. • Dynamic Programming • Game Theory • Queuing Theory •

One final note: The techniques and methods discussed herein can be superb tools. However there are no replacements for the ultimate managerial tool, “Common Sense.6”

5Marshal McLuhan, Understanding Media: The Extensions of Man, 1964.6com·mon sense (n)—Sound practical judgment derived from experience rather than study; Encarta® World English Dictionary ©1999 Microsoft Corporation. All rights reserved. Developed for Microsoft by Bloomsbury Publishing Plc


PRoBLeM DeFINItIoN

Occam’s Razor: one should not increase beyond what is necessary the number of entities required to explain anything7.

In other words, the simplest solution is frequently the correct one.

You’re driving along Interstate 25 north of the city of Cheyenne, Wyoming. Suddenly, you hear that annoying sound. It sounds like an airplane coming in for a landing. You get a sick feeling in the pit of your stomach. You have a flat tire. For those of you who have never been on Interstate 25 north of Cheyenne Wyoming, with the exception of a few antelope and scrub brush, there’s no there, there. You pull over and check—yes, the tire is flat, and what do you do? What exactly is the problem that you have to solve? The obvious answer is:

“Hey stupid, I have a flat!” Let’s examine the situation. We’ll get back to our tire later.

Flomberg’s Law #2: You can’t make a decisionif you don’t identify the problem

Before the problem can be analyzed and a decision made, you must be aware that a problem exists. In the tire example above, your problem is that you’re stuck on the side of the road with a flat tire, one hundred miles from anywhere. Or is that the problem?

Frequently the apparent problem is not the real problem but is problematic, or a symptom. A symptom indicates a problem exists, however it is not the prob-lem. Solving the symptom might feel good for a while, however the problem still exists. The successful manager must distinguish between the problem and the symptoms. Treating a gunshot wound with pain medication only addresses a symptom of the situation—it might alleviate the pain. It does not heal the gunshot.

7Occam’s Razor is a logical principle attributed to the 14th Century philosopher William of Occam

Section One 15

Another situation that the manager needs to deal with is the bias built into the word, problem. The dictionary definition of the word problem is: a question or puzzle that needs to be solved. A problem is not necessarily a negative situation. A problem simply means that a decision must be made. If there is no decision to be made there, is no problem—only a course of action. Problems can be either positive or negative. An opportunity can present itself that has choices. The selection of the choice can affect the return. The flat tire situation only indicates that there might be a problem. In the flat tire situation, we can assume that the problem is a negative one. But is the flat a problem? It is not. The flat is a state of nature. It exists. Deal with it. “I have been selected to three colleges.” Is that a problem? So, what is a problem? Is there a problem? The problem is, “Which College do I go to?” When identifying a problem, you must also deal with bias: “We’ve always done it that way!” This bias may take the form of the timeworn statements like: “We have to live with that problem. It’s not worth fixing.”

A Real Live Problem (or Is It a Problem?)

Many years ago when I was a computer programmer, I was working on a cus-tomer’s site, performing maintenance on the company’s general ledger software. When I was looking for test data, my team leader, who was an employee of the company, pointed me to a file. As I was going through that file I realized that I was actually in the real corporate ledger. Did they want me poking around in there? Then when I started testing my software changes, I found that the books balanced to a $.23 difference. When I inquired, I was told that the, “problem had existed for a while, no one wanted to spend the time to resolve it.”

Was there a problem? Were there two problems? Perhaps there were no prob-lems? Management obviously felt there were no problems. Is bias a problem? Does the fact that I was representing a trusted consulting firm enter into the situation? If you were the auditor, what would you say about this situation? I’ll come back to this in a moment.

We can list some of the properties of problem definition. The list might contain:

who is the person who is making the decision (the Decision Maker)? At what level does the problem have to be solved? Can a team lead


solve the problem or does it need to be escalated to the CEO? Perhaps some level of authority between the two is adequate. the Goal—What are you trying to do? What is the Big Picture? Do you want to maximize profit? Do you want to minimize cost? Constraints—What are the constraints that influence the deci-sion? They can be monetary, legal, geographical, ethical or even governmental. Alternatives—What are the alternatives for you to consider? The alternatives are frequently mutually exclusive.Payoffs or Penalties—Each alternative should have payoffs or penal-ties. What are they? Probabilities—Review past situations. Can history present probabili-ties of success or failure?

CASe StUDIeS

Flomberg’s Law #3: Doing nothing is always a valid alternative

You are a geographer and you must compute the approximate circumfer-1. ence of the world. You know the distance from Boston to Los Angeles is approximately 3000 miles. Using commonly available information, deduce a method for computing the circumference of the world. The method is the goal of this exercise, not the answer.In the Flat Tire situation, what are the possible actions? What is the prob-2. lem? What if you are 75 years old and have a bad heart—how does that affect the situation?In the “Real Life Problem” example, identify the problem(s)3. In 1962, Rachel Carlson wrote the book 4. Silent Spring. This book is consid-ered by many to be the beginning of the Ecology movement. You work for the Secretary of the Interior. He has read the book and has assigned you the responsibility of “Fixing the Problem.” Identify the problem. In 1911, a factory in New York City called, “The Triangle Shirtwaist 5. Company,” burned down. One hundred forty-six of the 500 people who worked there died. You work for the Mayor of New York. He called you into his office and told you to “Fix the Problem. Make sure that it never happens again” Identify the problem.

Section One 17

In 1965, Ralph Nader kicked off his career by publishing the book6. Unsafe at any Speed. This book killed the Chevrolet Corvair. You have been given an advance copy of the book and told by your boss, the CEO of GM to “Fix the Problem.” Identify the problem.You work for a major financial company. Your boss has been watching 7. the giants crumble and fall. He is nervous about spending time in “Club Fed” 8 for any oversight he might have missed. He has called you into his office and asked you to see if there is a problem, and if there is: “Fix the Problem.” Is there a problem? Why or Why not? If there is a problem, identify it. What information do you need in order to establish what the problem is?Your Local Sausage Manufacturing Company is being hit with a series of 8. complaints of sausages exploding when they are thawed out. Using the six steps listed below identify the problem.

Define the problem a. Identify alternative solutions b. Identify future outcomes c. Identify future returns, positive or negative d. Select and implement a modeling techniquee. Decidef.

management decision theory

Documents